User:Charles Matthews/Painstakingness

Wikipedia doesn't stay the same: it changes. Not only does the content grow, and not only does the editing community renew itself, but the nature of the actual task shifts about. This phenomenon can be expressed in anxious language of “backlogs”, or managerial shorthand about where people could be assigned (if there were any volunteers willing to go where they are told, an unlikely story). The real point is about priorities, given that the large editing community is also limited to a few thousand individuals, and their view of what requires doing determines what will happen.

Priorities are matched by some sort of narrative: what Wikipedia set out to do and be, what turned out to be easy to achieve and what remained elusive in terms of content and the construction of a group of committed people who would put together articles and policy and a manual, administration and technical support, and ultimately fundraising and public relations. Given the volume of criticism and comment, it is not so simple to achieve any sort of perspective on what has been done so far. Let alone a reconciliation of the multiple perspectives that are offered in many forums. But we cannot wait forever for the return of whatever jury it is that is going to offer a verdict on the site and its work (better, the WMF’s global package of sites and their impact). If the point is to use a narrative as a personal guide on how best to contribute, complex history is not as useful as an arc of story explaining what worked, what failed, and what was learned. If the question is “how to improve Wikipedia today?”, a simple but informed answer should carry more weight than a tangled response that still carries a burden of the past as baggage, not as what has been learned.

We learned first how to grow: this took Wikipedia to 2005. This growth was in the content, by quantitative measures. It was also in traffic, powered by the integrated use of hypertext (wikify on sight). Then content policy was supplemented by qualitative guidelines designed to make content better. This takes us to about 2008. At that point, say seven years in, we had a fairly good idea of what Wikipedia articles could achieve at best: not flashy in writing style, they could respectably collate much of the best information and views on topics over a vast range of topics (unattempted yet by encyclopedias). This is the point at which the “first draft” becomes the appropriate metaphor. If our site were a book, which it is not, the point was reached at which the scope of what should be attempted in the project could at least be discussed by addressing what was already there. The FA model, namely a treatment like a short book chapter with thorough footnoting and readable prose, with comprehensive factual content and illustrations, could be seen (in 1 article per 1000); the topics included marked out areas for such treatment that arguably could be written up in such a way, if not on any timescale that made practical sense.

The point of this essay, though, is to ask and answer the question: and then what? Some potential answers are rejected: it makes no sense, looking at the reference material that is out there, to say that “no more topics to add” has cramped the project. If you only research using a search engine, that is still nothing like true, given that new topics arise, but also that more and more material is added to the Web, as the years go by. The argument that quality does not always increase monotonically (poor editing and outright vandalism can damage articles, giving a downtick to quality) doesn’t prove that the general quality cannot be brought up by steady work, because it certainly can. The coarse numbers of edits, or accounts, or editors, have little to do with the actual productivity of the community. That’s because “working smart” can add and improve content with little fuss. Some 80/20 laws probably apply here as elsewhere: to understand what really pushes the project ahead, you have to look (roughly speaking) at the 20% of editors who add content rather than massage it, and the 20% of their edits that are big improvements.

So, back to the narrative. What has been going on the in past couple of years in the English Wikipedia? We have discovered, have we not, the virtues of painstakingness in relation to existing and added content. This brings the site and its ethos more into line with the values of academia. Not, it must be said, by the rather pompous route of “respect for expertise” or credentialism. As some of the more easy-to-research areas look more complete, the difficulties of writing good reference material more as a pioneer than as a latecomer come into focus for editors. “How do you know that?” as a skeptical challenge does become the more reflective “how do I know that?” for any editor working outside the comfort zone of familiar topics on which we tend to start, but also to outgrow. “What are the reference sources I need?” is a question that typically has answers (known, naturally, to experts), but the answers need not remain arcane for ever. Much that does need still to be added to the site fits with existing coverage, and checking exactly how can be a good education in some subtle points of understanding, revealing both misconceptions in the editor and flaws in what is already there where apparent inconsistencies occur. It is not foolish to say that working on Wikipedia can enable a dedicated researcher to find mistakes in other reference sources and scholarly works: it is an apt comment on where we are, in pushing out into territory that is less familiar and accessible.